J. K. Evans' well-documented article, ‘Wheat production and its social consequences in the Roman world’, correctly makes the point that ‘the evidence with regard to wheat yields is at once meagre and plainly contradictory’. The difficulty in assessing yields arises, of course, from the character of the available source material; namely, literary sources. The information comes from the hands of men such as Cicero and Varro who were concerned with matters other than specific data on the cultivation and production of grains, and who probably never sowed or reaped a modius of wheat. What was lacking until recently was a bona-fide document from the hands of a farmer or a community intimately concerned with the growing of wheat. We now have one such document, P. Colt 82 of the seventh century A.D., that fills a gap in the evidence for yields for both wheat and barley.